Diabetes mellitus is a major cause of morbidity and mortality in the United States, and individuals with impaired glucose tolerance (IGT) or a history of gestational diabetes (GDM) are at increased risk of developing this disease. Research to test whether it is possible to prevent the occurrence of NIDDM is a national priority. The University of Pittsburgh proposes a primary prevention trial that optimally tests the hypothesis that NIDDM is preventable through weight loss and physical activity and simultaneously addresses whether an intervention that is more easily adopted by the health care system is of benefit. Targeted advertisement and chart review will be used to identify 100 subjects with IGT (confirmed on repeat OGTT and with at least one 2-hour BS>170); 100 GDM subjects will be identified from a registry of over 1000 women with GDM; we aim for 40% of participants to be African American. These 200 high risk subjects will be randomized to a control group, a standard life-style intervention (SLI), or an enhanced life-style intervention (ELSI). The goals for the two intervention groups will be a 10% weight loss and 3 hours/week of physical activity. To achieve these goals, the SLI group will participate in a standard diet/exercise program that could be easily implemented by the health system; community-led exercise groups will also be available. The ELSI group will be given enhanced treatment to optimally test the prevention hypothesis; these subjects will be seen individually on a monthly basis; intervention will be prescribed as needed to achieve and maintain the weight and exercise goal. The primary outcome will be development of NIDDM (WHO criteria; confirmed by second test). Other endpoints will be changes in glucose, insulin resistance, insulin secretion, and the cardiovascular risk profile. During screening, some subjects will be identified as having non-fasting diabetes (NFD; FBS <140; 2 hour BS>200). These subjects will participate in a secondary prevention trial using a 2 x 2 factorial design with the two life-style conditions (SLI vs ESLI) crossed with 2 drug conditions (metformin vs placebo). The primary outcome measures for this secondary prevention study will be development of fasting hyperglycemia and incidence or progression of diabetic complications. The University of Pittsburgh is well qualified to participate in this trial as we have extensive experience in life-style interventions, exercise, clinical trials, diabetic complications, and in the medical management of diabetes.